Skip to content

Development Log.md

Latest commit

 

History

History
75 lines (58 loc) · 3.07 KB

Development Log.md

File metadata and controls

75 lines (58 loc) · 3.07 KB
Raw

2024/12/03

  • Try to review what I have done
  • The implementation of the projected density of states
  • Basic structure of the MemoryElectronicFriction
  • Learn the way to turn MemoryElectronicFriction into a package
  • Do something similar with NQCDModel
  • Turn MemoryElectronicFriction into a semi-package by defining a lot of module
  • Create the type of model for MemoryElectronicFriction
  • Call the DOS function by MemoryElectronicFriction module
  • (Top level function defined in MemoryElectronicFriction.jl module)

2024/12/05

  • Build the module Distributions Lorentzian and FermiDirac
  • Coordinate module Distributions with Parent module MemoryElectronicFriction

2024/12/09

  • Implemented the bath discretization from NQCModels.jl
  • Try to do the subtype graph for the MemoryElectronicFriction. 人都癫 I hate this crap.

2024/12/10

  • Implement the $\Gamma(\omega_1,\omega_2)$ on MemoryElectronicFriction (Not done yet, but making progress on dev_Gamma.jl)
  • bath.bathstates Assume our electronic distribution is following the Fermi Dirac distribution? 100 electron 200 states groundstate
  • Implement $V_{ak}(\omega)'$ onto MemoryElectronicFriction
  • If I have time, try to make the plot like alex did.
  • Have a look into those different package in julia and compare to Alex's code.
  • Changed the name of Distributions into DistributionTools
  • Added a gaussian struct into DistributionTools

2024/12/11

  • Have a look into the Cauchy principle value. Have a think about the implementation.
  • Try to Implement the $\Gamma(\omega_1,\omega_2)$ The configuration index summation.
  • Try to figure out concepts of $R_{ak}$ ? Is it okay to have $$ R_{ak} = R_{ka}^*? $$

I assume it should be equal because the symmetry?

Reason: The coupling terms in our Hamiltonian is identical if we consider them are real value.

In our case, it can boil down to the $$ G_{a}^{\text{ret}}(\omega)T_kG_{k}^{\text{ret}(0)}(\omega) ?= G_{k}^{\text{ret}}(\omega)T_aG_{a}^{\text{ret}(0)}(\omega) $$ In our discussion above we should have $$ T_a = T_k $$ Because both of them the coupling between impurity and bath. Should be identical. Now we need to figure out $$ G_a^{ret} G_k^{ret(0)} = G_k^{ret}G^{ret(0)}_a ? $$

  • $R_ak$ fully understood!!!!!!!!!!

2025/1/14

  • Review the structure(Julia hierarchy module) of the MemoryElectronicFriction
  • Make a schematic about coding the R_{kk'}
  • Make HokseonPlots.jl for my own usage

2025/6/29

  • Review the structure of MemoryElectronicFriction

2025/7/1

  • module AndersonImpurityFrictions built with submodule ImaginaryGreens where has Raa and Rak (see (A48) and (A47) in brandbyge paper)
  • Renamed module BrandbygeModel into AndersonImpurityModels. Updated the type of BrandbygeAdsorbate <: AndersonImpurityModel
  • Set up CI tests for BrandbygeAdsorbate, Raa and Rak.
  • Included dev version of NQCModels into Project.toml for MemoryElectronicFriction